For more than 800,000 Americans each year, unrelieved cancer pain is a major health problem attributed in part to assessment difficulties and to lack of knowledge about using the analgesics known to be effective for 85 percent to 95 percent of cancer patients with pain. As a solution to the unnecessary suffering, we propose a randomized clinical trial to compare effects of usual-care in the UW Cancer Center or a protocol of computerized pain assessment and provider decision support on: a) patient outcomes (pain intensity, quality, relief; satisfaction with pain level) in a diverse sample of 176 patients with pain related to bone metastasis; b) provider outcomes (pain documentation; appropriateness of prescribed analgesics; clinic visit length) in a sample of oncology specialists; and c) patient-provider partnership outcomes (patient-centered discussion content during audiotaped clinic visits). In a pre-test and post-test design, patients will complete usual pain assessments and use a new computerized pain assessment tool to document their pain. Pain data from patients assigned to the new protocol will be shared with their oncologists along with a computer generated consultation that includes a list of analgesics appropriate for the character of the patient's pain. The interaction between the patient and the provider during routine clinic visits will be audiotaped and analyzed for types of communication during the visit. One week later, patients will complete the computerized pain assessment tool to document the effect of any therapies prescribed for their pain. Additionally we will describe patient and provider receptivity to the computerized pain tool using semi-structured questions and applying content analysis to the data. Study findings will be used to guide revisions of the tool and to plan organizational level implementation of this new technology as an interface with the electronic health care record.